论文部分内容阅读
目的采用偏最小二乘(PLS)回归算法分析耐甲氧西林金黄色葡萄球菌(MRSA)检出率与抗菌药物使用的相关性,为指导抗菌药物的合理使用提供依据。方法以抗菌药物的用药频度(DDDs)为自变量,以同期及滞后不同时期MRSA的检出率为因变量,建立PLS回归模型。根据各抗菌药物的回归系数,筛选与MRSA检出率相关性较显著的品种,明确不同品种对MRSA检出率的贡献度。结果 MRSA的检出率与头孢地秦、克拉霉素的DDDs呈显著正相关,与氟康唑的DDDs呈显著负相关。头孢地秦、克拉霉素、氟康唑的PLS回归系数分别为0.119、0.119、-0.115。结论本研究证明了PLS回归算法在MRSA检出率与抗菌药物使用相关性研究中的有效性,可为病原菌耐药性研究提供新的思路。
Objective To analyze the correlation between the detection rate of methicillin-resistant Staphylococcus aureus (MRSA) and the use of antimicrobial agents by partial least squares (PLS) regression algorithm and provide the basis for the rational use of antimicrobial agents. Methods Taking the frequency of antimicrobial medication (DDDs) as independent variable and the detection rate of MRSA in different periods of the same period and lag as the dependent variable, a PLS regression model was established. According to the regression coefficients of each antibacterial, the strains with significant correlation with the detection rate of MRSA were screened, and the contribution of different varieties to the detection rate of MRSA was clarified. Results The detection rate of MRSA was positively correlated with DDDs of cefodizime and clarithromycin, but negatively correlated with DDDs of fluconazole. The PLS regression coefficients of cefodizime, clarithromycin and fluconazole were 0.119, 0.191 and -0.115, respectively. Conclusion This study demonstrates the validity of the PLS regression algorithm in the study of the correlation between the detection rate of MRSA and the use of antimicrobial agents, which may provide new ideas for the study of drug resistance of pathogens.